Biological material component analysis method and system based on combination of Internet of Things and neural network

Abstract

The invention relates to the technical field of the Internet of Things, and discloses a biological material component analysis method and system based on the combination of the Internet of Things and a neural network, and the method comprises the steps: obtaining material sample data of a to-be-detected biological material and environment state data of an environment where the to-be-detected biological material is located, and extracting an intrinsic attribute index of the to-be-detected biological material from the material sample data, calculating the basic substance content of the biological material to be detected; calculating the characteristic absorption rate of the to-be-detected biological material so as to analyze the characteristic absorption intensity of the to-be-detected biological material; constructing a spectral line graph corresponding to the to-be-detected biological material, and re-extracting fingerprint spectral information of the to-be-detected biological material from the spectral line graph; and inputting the fingerprint spectrum information into a biological detection unit of a preset Internet of Things, outputting a corresponding induction signal sequence by using the biological detection unit so as to identify molecular configuration information of the to-be-detected biological material, and analyzing the material composition of the to-be-detected biological material by using a trained neural network. The accuracy of biological material component analysis can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of biomaterial analysis, and in particular to a biomaterial component analysis method and system based on the Internet of Things combined with a neural network. Background Art

[0002] In the fields of biomaterial research, pharmaceutical development, food testing, etc., biomaterial component analysis is a key link in controlling material quality and ensuring application safety. For example, in food testing, it is necessary to identify the components of biomaterials in food raw materials to prevent harmful components from affecting food safety.

[0003] Existing methods for analyzing the composition of biomaterials mainly include high-performance liquid chromatography, gas chromatography, etc. These methods require complex pretreatment operations on biomaterial samples and rely on large-scale precision instruments. Not only is the analysis cycle long, but it is also difficult to achieve real-time dynamic analysis of biomaterials. At the same time, they are not combined with Internet technology, which leads to inaccurate detection and analysis of biomaterial mixtures with complex components or target components with extremely low content, thereby reducing the accuracy of biomaterial composition analysis. Summary of the Invention

[0004] The present invention provides a biomaterial component analysis method and system based on the Internet of Things combined with a neural network, the main purpose of which is to improve the accuracy of biomaterial component analysis.

[0005] To achieve the above objectives, the present invention provides a method for analyzing biomaterial components based on the Internet of Things combined with a neural network, comprising:

[0006] Obtaining material sample data of the biomaterial to be tested and environmental status data of its environment, extracting intrinsic property indicators of the biomaterial to be tested from the material sample data to calculate the basic substance content of the biomaterial to be tested;

[0007] Scanning the biomaterial to be tested using a preset multi-band optical probe, collecting input optical signals and transmitted optical signals during the scanning process, and calculating the characteristic absorption rate of the biomaterial to be tested based on the input optical signal, the transmitted optical signal, and the basic substance content to analyze the characteristic absorption intensity of the biomaterial to be tested;

[0008] constructing a spectrum line graph corresponding to the biomaterial to be tested according to the characteristic absorption intensity, and re-extracting fingerprint spectrum information of the biomaterial to be tested from the spectrum line graph;

[0009] The fingerprint spectrum information is input into a biological detection unit of a preset Internet of Things, and the biological detection unit is used to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be tested. The molecular configuration information is combined with the fingerprint spectrum information, and the material composition of the biological material to be tested is analyzed using a trained neural network.

[0010] Optionally, extracting the intrinsic property index of the biomaterial to be tested from the material sample data includes:

[0011] performing interference removal processing on the material sample data to obtain clean sample data;

[0012] Identifying a feature identifier corresponding to the clean sample data, and calculating an identifier correlation degree between the feature identifier and a preset feature identifier;

[0013] extracting material characteristic data from the clean sample data according to the identification correlation degree;

[0014] The material characteristic data is analyzed to obtain the intrinsic property index of the biomaterial to be tested.

[0015] Optionally, calculating the identification correlation between the feature identification and a preset feature identification includes:

[0016] Performing vectorization processing on the feature identifier and the preset feature identifier respectively to obtain a first feature vector and a second feature vector;

[0017] Performing dimensionality reduction processing on the first eigenvector and the second eigenvector respectively to obtain a first reduced dimensionality vector and a second reduced dimensionality vector;

[0018] The first dimensionality reduction vector and the second dimensionality reduction vector are combined to calculate the identification correlation between the feature identifier and the preset feature identifier using the following formula:

[0019]

[0020] Among them, A represents the identification correlation between the feature identification and the preset feature identification, Represents the first dimensionality reduction vector corresponding to the ath identifier in the feature identifier, Represents the second dimensionality reduction vector corresponding to the ath identifier in the preset feature identifier, Indicates the maximum value between two values.

[0021] Optionally, the extracting the intrinsic property index of the biomaterial to be tested from the material sample data to calculate the basic substance content of the biomaterial to be tested further includes:

[0022] Extracting core attribute dimensions from the intrinsic attribute indicators, the core attribute dimensions including: structural feature dimensions, component feature dimensions, and environmental response feature dimensions;

[0023] Standardizing the structural characteristic dimension, the component characteristic dimension, and the environmental response characteristic dimension respectively to obtain a structural standard value, a component standard value, and a response standard value;

[0024] The basic substance content of the biomaterial to be tested is calculated using the following formula by combining the structure standard value, the component standard value, and the response standard value:

[0025]

[0026] Wherein, E represents the basic substance content of the biological material to be tested, Indicates the standard value of the structure, Indicates the standard value of the component, Indicates the response standard value, Represents the dimensionality correction factor.

[0027] Optionally, the calculating the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal and the basic substance content includes:

[0028] Measuring the detection distance between the biological material to be tested and the multi-band optical probe;

[0029] determining a propagation distance of the optical signal of the biomaterial to be tested based on the detection distance;

[0030] Measuring the signal intensities of the input optical signal and the transmitted optical signal respectively to obtain the input optical signal intensity and the projected optical signal intensity;

[0031] The characteristic absorption rate of the biomaterial to be tested is calculated using the following formula based on the propagation distance, the input optical signal intensity, the projected optical signal intensity, and the basic substance content:

[0032]

[0033] in, represents the characteristic absorption rate of the biological material to be tested, d represents the propagation distance, Indicates the input optical signal strength, It represents the intensity of the projected optical signal, and E represents the basic substance content.

[0034] Optionally, the analyzing the characteristic absorption intensity of the biomaterial to be tested further comprises:

[0035] querying a probe attribute parameter of the multi-band optical probe, determining a light signal band of a light signal of the biological material to be measured based on the probe attribute parameter;

[0036] measuring a molecular vibration frequency of the biological material to be measured, and calculating a material characteristic coefficient of the biological material to be measured based on the molecular vibration frequency;

[0037] combining the material characteristic coefficient, the light signal band, and the characteristic absorption rate, and calculating an absorption contribution value of the biological material to be measured by using the following formula:

[0038]

[0039] wherein, the absorption contribution value of the biological material to be measured, the characteristic absorption rate, k represents the material characteristic coefficient, the light signal band;

[0040] analyzing a characteristic absorption intensity of the biological material to be measured based on the absorption contribution value.

[0041] Optionally, the extracting the fingerprint spectrum information of the biological material to be measured from the optical spectrum diagram comprises:

[0042] performing background signal elimination processing on the optical spectrum diagram to obtain a background-corrected spectrum;

[0043] performing filtering processing on the background-corrected spectrum to obtain a filtered spectrum;

[0044] performing feature enhancement processing on the filtered spectrum to obtain an enhanced spectrum;

[0045] identifying a characteristic absorption region in the enhanced spectrum, and extracting an absorption mode corresponding to the characteristic absorption region;

[0046] determining fingerprint spectrum information corresponding to the biological material to be measured according to the absorption mode.

[0047] Optionally, the extracting the absorption mode corresponding to the characteristic absorption region comprises:

[0048] calculating an integral intensity corresponding to each region in the characteristic absorption region, and determining an intensity distribution feature in the integral intensity;

[0049] locating a center band corresponding to the characteristic absorption region according to the intensity distribution feature;

[0050] calculating a half-peak width corresponding to each region in the characteristic absorption region, and analyzing a contour shape corresponding to the characteristic absorption region;

[0051] The geometric properties corresponding to the contour shape are analyzed, and the absorption pattern corresponding to the characteristic absorption region is generated by combining the central band, the half-peak width and the geometric properties.

[0052] Optionally, the utilizing the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be detected includes:

[0053] Analyzing data features in the fingerprint spectrum information, and configuring a combination of sensing elements in the biometric detection unit according to the data features;

[0054] Adjusting the sensitivity of the sensing element combination to obtain an optimized sensing combination;

[0055] Performing characteristic response processing on the fingerprint spectrum information using the optimized sensing combination to obtain a characteristic response spectrum;

[0056] Converting the characteristic response spectrum into a time-series electrical signal, and generating the sensing signal sequence according to the time-series electrical signal;

[0057] Response pattern features are extracted from the sensing signal sequence, and molecular configuration information of the biomaterial to be tested is identified based on the response pattern features.

[0058] In order to solve the above problems, the present invention also provides a biomaterial component analysis system based on the Internet of Things combined with a neural network, the system comprising:

[0059] a substance content calculation module, configured to obtain material sample data of the biomaterial to be tested and environmental status data of the environment in which it is located, extract the intrinsic property index of the biomaterial to be tested from the material sample data, and calculate the basic substance content of the biomaterial to be tested;

[0060] an absorption intensity analysis module, configured to scan the biomaterial to be tested using a preset multi-band optical probe, collect input optical signals and transmitted optical signals during the scanning process, and calculate the characteristic absorption rate of the biomaterial to be tested based on the input optical signal, the transmitted optical signal, and the basic substance content, so as to analyze the characteristic absorption intensity of the biomaterial to be tested;

[0061] a spectral information extraction module, configured to construct a spectral line graph corresponding to the biomaterial to be tested based on the characteristic absorption intensity, and re-extract the fingerprint spectral information of the biomaterial to be tested from the spectral line graph;

[0062] The biomaterial analysis module is used to input the fingerprint spectrum information into a biological detection unit of a preset Internet of Things, use the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biomaterial to be tested, combine the molecular configuration information with the fingerprint spectrum information, and use a trained neural network to analyze the material composition of the biomaterial to be tested.

[0063] Compared to the problems described in the background art, the present invention extracts the intrinsic property indicators of the biomaterial to be tested from the material sample data, thereby obtaining key characteristic information of the biomaterial to be tested, providing data support for subsequent basic substance content analysis and processing. Furthermore, the present invention calculates the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal, and the basic substance content, which can reflect the absorption characteristics of the biomaterial to be tested for light in different bands, thereby providing a basis for analyzing the characteristic absorption intensity. Furthermore, the present invention constructs a spectral line graph corresponding to the biomaterial to be tested based on the characteristic absorption intensity, which can intuitively display the absorption characteristic distribution of the biomaterial to be tested in different bands. The spectral line graph can improve the integrity of subsequent fingerprint spectrum information extraction. Furthermore, the present invention inputs the fingerprint spectrum information into a biological detection unit preset in the Internet of Things, thereby realizing remote transmission and real-time processing of detection data. The corresponding sensing signal sequence output by the biological detection unit provides a data basis for the subsequent identification of molecular configuration information. Therefore, the method and system for biomaterial component analysis based on the Internet of Things combined with neural networks provided in the embodiments of the present invention can improve the accuracy of biomaterial component analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A schematic diagram of a process for analyzing biomaterial components based on the Internet of Things combined with a neural network, according to one embodiment of the present invention;

[0065] Figure 2 A schematic diagram of the training process in the method for analyzing biomaterial components based on the Internet of Things combined with a neural network provided by the present invention;

[0066] Figure 3 A schematic diagram of modules for implementing a biomaterial component analysis system based on the Internet of Things combined with a neural network, provided for one embodiment of the present invention, wherein 201 is a trigger condition analysis module, 202 is a transportation timeliness accuracy calculation module, 203 is a route matching coefficient calculation module, 204 is a scheduling fitness analysis module, and 205 is a logistics route scheduling optimization module.

[0067] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0068] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0069] The embodiments of the present application provide a method for analyzing the composition of biomaterials based on the Internet of Things combined with a neural network. The execution subject of the method for analyzing the composition of biomaterials based on the Internet of Things combined with a neural network includes, but is not limited to, at least one of electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiments of the present application. In other words, the method for analyzing the composition of biomaterials based on the Internet of Things combined with a neural network can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc.

[0070] Reference Figure 1 FIG. 1 is a flow chart of a method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to an embodiment of the present invention. In this embodiment, the method for analyzing biomaterial components based on the Internet of Things combined with a neural network includes:

[0071] S1. Obtain material sample data of a biomaterial to be tested and environmental status data of its surroundings, extract intrinsic property indicators of the biomaterial to be tested from the material sample data, and calculate the basic substance content of the biomaterial to be tested.

[0072] The present invention can obtain key characteristic information of the biomaterial to be tested by extracting the intrinsic property indicators of the biomaterial to be tested from the material sample data, and provide data support for subsequent basic substance content analysis and processing.

[0073] Among them, the biological material to be tested is a biological sample that needs to be identified for its components, the material sample data is the original record information related to the biological material to be tested, the environmental status data includes the temperature, humidity and lighting conditions of the environment in which the biological material to be tested is located, the intrinsic property index is used to describe the physical and chemical properties of the biological material to be tested, and the basic substance content represents the relative proportion of the main components in the biological material to be tested. If the biological material is animal muscle tissue, the basic substance content includes the mass ratio of protein and cellulose, the percentage of water in the total mass of the material, etc.

[0074] As an embodiment of the present invention, extracting the intrinsic property index of the biomaterial to be tested from the material sample data includes:

[0075] performing interference removal processing on the material sample data to obtain clean sample data;

[0076] Identifying a feature identifier corresponding to the clean sample data, and calculating an identifier correlation degree between the feature identifier and a preset feature identifier;

[0077] extracting material characteristic data from the clean sample data according to the identification correlation degree;

[0078] The material characteristic data is analyzed to obtain the intrinsic property index of the biomaterial to be tested.

[0079] The clean sample data is the material sample data after removing interference, the feature identifier is a marker used to characterize the specific attributes of the clean sample data, the preset feature identifier is a predefined typical feature identifier of the biological material, the identifier correlation degree is used to measure the correlation strength between the feature identifier and the preset feature identifier, and the material characteristic data is the part of the clean sample data related to the biological material characteristics.

[0080] Optionally, interference removal processing of the material sample data can be achieved through a wavelet transform method, and identifying feature identifiers corresponding to the clean sample data can be completed using feature extraction tools, such as OpenCV, Scikit-learn and other library functions. The correlation degree is compared with a set threshold. The set threshold can be 0.75, and can also be adjusted according to specific applications. If the correlation degree is higher than the set threshold, the material characteristic data is extracted from the clean sample data through a selection function, and the selection function includes the pandas.DataFrame.iloc[] method. Index analysis of the material characteristic data can be achieved through a parsing program, and the parsing program is written in Python.

[0081] Furthermore, as an optional embodiment of the present invention, the calculating the identification correlation between the feature identification and the preset feature identification includes:

[0082] Performing vectorization processing on the feature identifier and the preset feature identifier respectively to obtain a first feature vector and a second feature vector;

[0083] Performing dimensionality reduction processing on the first eigenvector and the second eigenvector respectively to obtain a first reduced dimensionality vector and a second reduced dimensionality vector;

[0084] In combination with the first dimensionality reduction vector and the second dimensionality reduction vector, the identification association degree between the feature identifier and the preset feature identifier is calculated using the following formula.

[0085] The first feature vector is a vector representation formed after numerical conversion of the feature identifier, the second feature vector is a vector representation formed after the same conversion mode of the preset feature identifier, the first dimension reduction vector is a result obtained after dimension compression of the first feature vector, and the second dimension reduction vector is a result obtained after the same compression mode of the second feature vector.

[0086] Further, the vectorization processing of the feature identifier and the preset feature identifier can be realized by a word embedding model, such as a pre-training model such as Word2Vec or BERT, to convert the identifier text into a high-dimensional vector representation. The dimension reduction processing of the feature vector can be realized by a principal component analysis method to retain the main feature information in the vector while reducing the computational complexity. The calculated identifier correlation degree is compared with the set reference, which can be 0.75, or can be adjusted according to the specific application requirement. If the identifier correlation degree is higher than the set reference, it is considered that the feature identifier and the preset feature identifier have significant correlation.

[0087] Further, as another embodiment of the present application, the first dimension reduction vector and the second dimension reduction vector are combined, and the identifier correlation degree between the feature identifier and the preset feature identifier is calculated by using the following formula, which includes:

[0088]

[0089] Wherein A represents the identifier correlation degree between the feature identifier and the preset feature identifier, represents the first dimension reduction vector corresponding to the a-th identifier in the feature identifier, represents the second dimension reduction vector corresponding to the a-th identifier in the preset feature identifier, represents the maximum value between two numerical values.

[0090] The present application extracts the intrinsic attribute index of the biological material to be tested from the material sample data to calculate the basic substance content of the biological material to be tested, which can clearly indicate the quantity proportion of the main components in the biological material to be tested, and provide a basis for subsequent optical analysis. The basic substance content represents the relative proportion of the main components in the biological material to be tested.

[0091] As an embodiment of the present application, the extraction of the intrinsic attribute index of the biological material to be tested from the material sample data to calculate the basic substance content of the biological material to be tested further includes:

[0092] The core attribute dimension is extracted from the intrinsic attribute index, and the core attribute dimension includes a structure feature dimension, a component feature dimension and an environmental response feature dimension.

[0093] Standardizing the structural characteristic dimension, the component characteristic dimension, and the environmental response characteristic dimension respectively to obtain a structural standard value, a component standard value, and a response standard value;

[0094] The basic substance content of the biomaterial to be tested is calculated using the following formula in combination with the structure standard value, the component standard value and the response standard value.

[0095] Among them, the core attribute dimension is a key aspect reflecting the essential characteristics of the biomaterial to be tested, the structural characteristic dimension describes the microstructural characteristics of the material, the component characteristic dimension characterizes the basic constituent elements of the material, and the environmental response characteristic dimension reflects the reaction characteristics of the material to external conditions. The structural standard value is the value of the structural characteristic dimension after standardized conversion, the component standard value is the value of the component characteristic dimension after the same processing method, and the response standard value is the value of the environmental response characteristic dimension after the same processing method.

[0096] Furthermore, extracting core attribute dimensions from the intrinsic attribute indicators can be achieved through feature importance assessment methods, such as feature selection methods based on random forest or XGBoost algorithms. Standardization of the characteristic dimensions can be achieved through Z-score standardization methods to make data of different dimensions comparable. The calculated basic substance content is compared with a reference range, which can be pre-set based on the material type and application field. If the calculated result exceeds the reference range, it indicates that the material properties may be abnormal.

[0097] Furthermore, as another embodiment of the present invention, the basic substance content of the biomaterial to be tested is calculated using the following formula in combination with the structure standard value, the component standard value and the response standard value, including:

[0098]

[0099] Wherein, E represents the basic substance content of the biological material to be tested, Indicates the standard value of the structure, Indicates the standard value of the component, Indicates the response standard value, Represents the dimensionality correction factor.

[0100] S2. Scan the biomaterial to be tested using a preset multi-band optical probe, collect input optical signals and transmitted optical signals during the scanning process, and calculate the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal, and the basic substance content to analyze the characteristic absorption intensity of the biomaterial to be tested.

[0101] The present invention calculates the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal and the basic substance content, which can reflect the absorption characteristics of the biomaterial to be tested for light in different bands, and further provide a basis for analyzing the characteristic absorption intensity. The multi-band optical probe is a detection device that can emit optical signals of multiple wavelengths (such as ultraviolet, visible, and near-infrared bands), which can specifically capture the optical responses of different substances in the biomaterial. The input optical signal is the initial optical signal emitted by the multi-band optical probe, which can cover the characteristic absorption band of the target substance in the biomaterial to be tested. The transmitted optical signal is the optical signal remaining after the input optical signal passes through the biomaterial to be tested. The characteristic absorption rate represents the absorption ratio of the biomaterial to be tested for light in a specific band, and the characteristic absorption intensity reflects the strength of the biomaterial to be tested's ability to absorb light. Optionally, the collection of the input optical signal and the transmitted optical signal can be achieved by a photodetector, and the intensity is quantified by converting the optical signal into an electrical signal.

[0102] As an embodiment of the present invention, the step of calculating the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal, and the basic substance content includes:

[0103] Measuring the detection distance between the biological material to be tested and the multi-band optical probe;

[0104] determining a propagation distance of the optical signal of the biomaterial to be tested based on the detection distance;

[0105] Measuring the signal intensities of the input optical signal and the transmitted optical signal respectively to obtain the input optical signal intensity and the projected optical signal intensity;

[0106] The characteristic absorption rate of the biomaterial to be tested is calculated using the following formula in combination with the propagation distance, the input optical signal intensity, the projected optical signal intensity and the basic substance content.

[0107] Among them, the detection distance refers to the straight-line distance between the signal transmission port of the multi-band optical probe and the surface of the biomaterial to be tested, which is used to characterize the spatial interval between the starting point of the light signal transmission and the material; the propagation distance refers to the actual path length of the light signal from the transmission port, through the biomaterial to be tested, to the receiving port, and its value is related to the physical thickness of the material and the detection distance; the characteristic absorption rate is a quantitative indicator reflecting the ability of the biomaterial to be tested to absorb light in a specific band. The input optical signal intensity and the transmitted optical signal intensity are the intensity of the input optical signal before contacting the biomaterial to be tested and the intensity after passing through, respectively.

[0108] Further, the detection distance can be measured by an infrared distance measuring module, which is integrated with the multi-band optical probe. The module emits an 850 nm infrared light beam and receives the reflected signal to calculate the detection distance based on the triangulation principle. When determining the propagation distance, if the detection distance is less than or equal to 5 mm (i.e., the probe is close to the material surface), the propagation distance is equal to the actual thickness of the biological material to be measured (previously measured by an ultrasonic thickness gauge). If the detection distance is greater than 5 mm, the propagation distance is the sum of the material thickness and twice the detection distance (i.e., the distance of the light signal going back and forth between the probe and the material surface plus the distance through the material). When measuring the signal intensity, the ambient light interference needs to be removed by a filter circuit. The average value of the same band signal collected 10 times in succession is taken as the final intensity value to reduce the influence of random noise.

[0109] Further, as another embodiment of the present application, the propagation distance, the input optical signal intensity, the projected optical signal intensity, and the basic substance content are combined to calculate the characteristic absorption rate of the biological material to be measured using the following formula, which includes:

[0110]

[0111] wherein, represents the characteristic absorption rate of the biological material to be measured, d represents the propagation distance, represents the input optical signal intensity, represents the projected optical signal intensity, and E represents the basic substance content.

[0112] The present application calculates the characteristic absorption rate of the biological material to be measured to analyze the characteristic absorption intensity of the biological material to be measured, so that the strength and distribution of the light absorption ability of the biological material to be measured at different wave bands can be understood, thereby facilitating the construction of subsequent optical characteristic maps. The characteristic absorption intensity represents the light absorption ability of the biological material to be measured at a specific wave band.

[0113] As an embodiment of the present application, the analysis of the characteristic absorption intensity of the biological material to be measured further includes:

[0114] The probe attribute parameters of the multi-band optical probe are queried, and based on the probe attribute parameters, the light signal wave band of the light signal of the biological material to be measured is determined.

[0115] The molecular vibration frequency of the biological material to be measured is measured, and based on the molecular vibration frequency, the material characteristic coefficient of the biological material to be measured is calculated.

[0116] The material characteristic coefficient, the light signal wave band, and the characteristic absorption rate are combined to calculate the absorption contribution value of the biological material to be measured using the following formula:

[0117] Based on the absorption contribution value, a characteristic absorption intensity of the biological material to be measured is analyzed.

[0118] wherein, the light signal wave band refers to a wavelength range (unit: nm) covered by the light signal emitted by the multi-waveband optical probe, which is determined by the wave band configuration information in the probe attribute parameter, such as ultraviolet wave band (200-400 nm), visible wave band (400-760 nm), near-infrared wave band (760-2500 nm), different wave bands correspond to different characteristic absorption peaks of different substances in the biological material; the molecular vibration frequency is the inherent frequency of the chemical bond vibration in the molecule of the biological material to be measured, which reflects the specificity of the molecular structure, such as the amide bond vibration frequency of protein and the phosphodiester bond vibration frequency of nucleic acid; the material characteristic coefficient is a parameter (value range 0.9-1.1) quantifying the influence of the molecular vibration state on light absorption, and the calculation method is "material characteristic coefficient = measured value of molecular vibration frequency / standard value of molecular vibration frequency", the farther the coefficient deviates from 1, the greater the variation of the molecular structure, and the more significant the influence on light absorption; the absorption contribution value is a quantitative index of the comprehensive material characteristics, wave band characteristics and absorption rate, which reflects the actual absorption capacity of the biological material to light under specific conditions, and the greater the value, the stronger the absorption.

[0119] Further, the query probe attribute parameter can be realized by a parameter storage module provided by the device, which records the center wavelength, bandwidth, emission power and other information of each wave band of the probe, which can be directly read through the RS485 interface; the measurement of the molecular vibration frequency needs to use a confocal Raman spectrometer, the laser excitation wavelength is set to 532 nm, and the spectral resolution is better than 2 cm⁻¹; during the measurement, the biological material needs to be frozen sectioned (thickness 5-10 μm) to reduce the light scattering interference. Before calculating the material characteristic coefficient, the molecular vibration frequency data needs to be baseline corrected (using a polynomial fitting algorithm) and smoothed (the Gaussian filter window size is set to 7) to eliminate the influence of background noise.

[0120] Further, as another embodiment of the present application, the absorption contribution value of the biological material to be measured is calculated by using the following formula in combination with the material characteristic coefficient, the light signal wave band and the characteristic absorption rate:

[0121]

[0122] wherein, represents the absorption contribution value of the biological material to be measured, represents the characteristic absorption rate, k represents the material characteristic coefficient, represents the light signal wave band.

[0123] S3, according to the characteristic absorption intensity, a spectrum spectrum line graph corresponding to the biological material to be measured is constructed, and the fingerprint spectrum information of the biological material to be measured is extracted from the spectrum spectrum line graph.

[0124] The present invention constructs a spectral spectrum diagram corresponding to the biomaterial to be tested based on the characteristic absorption intensity, which can intuitively display the absorption characteristic distribution of the biomaterial to be tested in different bands. The spectral spectrum diagram can improve the integrity of subsequent fingerprint spectrum information extraction, wherein the spectral spectrum diagram is a continuous spectrum composed of the characteristic absorption intensity changes of the biomaterial to be tested in different bands, and the fingerprint spectrum information is the characteristic data in the spectral spectrum diagram that can uniquely characterize the characteristics of the biomaterial to be tested. Optionally, constructing the spectral spectrum diagram corresponding to the biomaterial to be tested can be achieved by a drawing program, such as the Matplotlib drawing library.

[0125] As an embodiment of the present invention, extracting the fingerprint spectrum information of the biological material to be tested from the spectrum line graph includes:

[0126] Performing background signal removal processing on the spectral line graph to obtain a background correction spectrum line;

[0127] Performing filtering on the background correction spectrum to obtain a filtered spectrum;

[0128] Performing feature enhancement processing on the filtered spectral lines to obtain enhanced spectral lines;

[0129] Identifying a characteristic absorption region in the enhanced spectrum and extracting an absorption mode corresponding to the characteristic absorption region;

[0130] According to the absorption pattern, fingerprint spectrum information corresponding to the biological material to be tested is determined.

[0131] Among them, the background correction spectrum line is the spectrum curve obtained after deducting non-specific background interference from the spectrum line diagram, the filtered spectrum line is the smooth curve obtained after suppressing the high-frequency noise in the background correction spectrum line, the enhanced spectrum line is the prominent characteristic curve obtained after amplifying the useful signal in the filtered spectrum line, the characteristic absorption area is the band interval in the enhanced spectrum line that exhibits significant absorption characteristics, and the absorption mode is the intensity distribution form presented by the characteristic absorption area.

[0132] Optionally, background subtraction processing of the spectral line graph can be achieved through a polynomial fitting method, filtering processing of the background correction spectrum line can be achieved through a moving average filtering method, feature enhancement processing of the filtered spectrum line can be achieved through a derivative spectroscopy method, and identification of characteristic absorption areas in the enhanced spectrum line can be achieved through a sliding window detection method.

[0133] Optionally, as an optional embodiment of the present invention, extracting the absorption pattern corresponding to the characteristic absorption region includes:

[0134] Calculating the integrated intensity corresponding to each region in the characteristic absorption region, and determining the intensity distribution characteristics in the integrated intensity;

[0135] Locating a central band corresponding to the characteristic absorption region according to the intensity distribution characteristics;

[0136] Calculate the half-maximum width corresponding to each region in the characteristic absorption region, and analyze the contour shape corresponding to the characteristic absorption region;

[0137] The geometric properties corresponding to the contour shape are analyzed, and the absorption pattern corresponding to the characteristic absorption region is generated by combining the central band, the half-peak width and the geometric properties.

[0138] Among them, the integrated intensity is the cumulative absorption intensity of the characteristic absorption area within a specific band range, the intensity distribution characteristic is the distribution law of the integrated intensity along the band, the central band is the band position with the maximum absorption intensity in the characteristic absorption area, the half-peak width is the band width corresponding to half of the maximum absorption intensity in the characteristic absorption area, the morphological method is a processing method for analyzing the shape of the spectral profile, and the geometric attribute is the mathematical representation corresponding to the profile shape, such as symmetry or steepness.

[0139] Optionally, the integral intensity corresponding to each area in the characteristic absorption area can be calculated by the trapezoidal integration method, and the intensity concentration interval of the characteristic absorption area can be obtained according to the intensity distribution law, so as to locate the central band corresponding to the characteristic absorption area. The half-peak width corresponding to each area in the characteristic absorption area can be calculated by the interpolation calculation method, and the geometric properties can be analyzed according to the indicators such as the eccentricity and compactness of the contour shape.

[0140] S4. Input the fingerprint spectrum information into a biological detection unit of a preset Internet of Things, use the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be tested, combine the molecular configuration information with the fingerprint spectrum information, and use a trained neural network to analyze the material composition of the biological material to be tested.

[0141] The present invention can realize remote transmission and real-time processing of detection data by inputting the fingerprint spectral information into a biological detection unit preset in the Internet of Things. The biological detection unit outputs a corresponding sensing signal sequence, which provides a data basis for the subsequent identification of molecular configuration information. The biological detection unit is a dedicated biometric detection component integrated into the Internet of Things system, which has data acquisition, conversion and transmission functions. The sensing signal sequence is a series of electrical signal responses generated by the biological detection unit based on the input fingerprint spectral information, reflecting the characteristic response pattern of the biological material to be tested.

[0142] As an embodiment of the present invention, the method of using the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be detected includes:

[0143] Analyzing data features in the fingerprint spectrum information, and configuring a combination of sensing elements in the biometric detection unit according to the data features;

[0144] Adjusting the sensitivity of the sensing element combination to obtain an optimized sensing combination;

[0145] Performing characteristic response processing on the fingerprint spectrum information using the optimized sensing combination to obtain a characteristic response spectrum;

[0146] Converting the characteristic response spectrum into a time-series electrical signal, and generating the sensing signal sequence according to the time-series electrical signal;

[0147] Response pattern features are extracted from the sensing signal sequence, and molecular configuration information of the biomaterial to be tested is identified based on the response pattern features.

[0148] Among them, the data feature is a discriminative quantitative indicator in the fingerprint spectral information, the sensing element combination is a set of elements in the biological detection unit for generating a response signal, the optimized sensing combination is an element combination that can better match the data feature after sensitivity adjustment, the characteristic response spectrum is the response distribution obtained after the optimized sensing combination processes the input information, the time series electrical signal is a set of electrical signals obtained by converting the characteristic response spectrum in time sequence, and the response pattern feature is a characteristic form that appears repeatedly or changes regularly in the sensing signal sequence.

[0149] Optionally, parsing the data features in the fingerprint spectrum information can be achieved through a feature extraction algorithm, and corresponding sensing elements are selected for combination configuration based on the intensity distribution and band characteristics in the data features. Sensitivity adjustment of the sensing element combination can be achieved through potentiometer adjustment or software gain control. Converting the characteristic response spectrum into a time-series electrical signal can be achieved through an analog-to-digital converter, and extracting response pattern features from the sensing signal sequence can be achieved through a time-frequency analysis method.

[0150] The present invention combines the molecular configuration information with the fingerprint spectrum information and uses a trained neural network to analyze the material composition of the biological material to be tested, thereby improving the accuracy and reliability of the material composition analysis, wherein the molecular configuration information is a representation of the spatial arrangement and structural characteristics of the molecules in the biological material to be tested.

[0151] Optionally, the trained neural network is an intelligent analysis network obtained by training with a large amount of biological sample data composed of known substances. The training process of the neural network includes: collecting molecular configuration information and fingerprint spectrum information of different biological materials as training samples, constructing a neural network architecture, setting the number of network layers and node parameters, iteratively training the network using training samples, and adjusting the network weights until the preset accuracy requirements are met. The neural network can accurately infer the material composition of unknown biological materials by learning the complex mapping relationship between molecular configuration information and fingerprint spectrum information. During analysis, the molecular configuration information and fingerprint spectrum information of the biological material to be tested are simultaneously input into the trained neural network, and the material composition results of the biological material to be tested are directly output through the forward propagation calculation of the network. Specifically, in order to further intuitively understand the lighting processing flow of the adaptive method of the intelligent lighting explosion-proof photosensitive controller in this application, you can refer to Figure 2 As shown in FIG, it is a schematic diagram of the training process in the method for analyzing the composition of biomaterials based on the Internet of Things combined with a neural network provided by the present invention. It should be noted that, in the present invention, Figure 2 The presented flowchart is only used for the training process of the biomaterial component analysis method based on the Internet of Things combined with a neural network, and is not limited to the training of the biomaterial component analysis method based on the Internet of Things combined with a neural network in different actual application scenarios.

[0152] Compared to the problems described in the background art, the present invention extracts the intrinsic property indicators of the biomaterial to be tested from the material sample data, thereby obtaining key characteristic information of the biomaterial to be tested, providing data support for subsequent basic substance content analysis and processing. Furthermore, the present invention calculates the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal, and the basic substance content, which can reflect the absorption characteristics of the biomaterial to be tested for light in different bands, thereby providing a basis for analyzing the characteristic absorption intensity. Furthermore, the present invention constructs a spectral line graph corresponding to the biomaterial to be tested based on the characteristic absorption intensity, which can intuitively display the absorption characteristic distribution of the biomaterial to be tested in different bands. The spectral line graph can improve the integrity of subsequent fingerprint spectrum information extraction. Furthermore, the present invention inputs the fingerprint spectrum information into a biological detection unit preset in the Internet of Things, thereby realizing remote transmission and real-time processing of detection data. The corresponding sensing signal sequence output by the biological detection unit provides a data basis for the subsequent identification of molecular configuration information. Therefore, the method and system for biomaterial component analysis based on the Internet of Things combined with neural networks provided in the embodiments of the present invention can improve the accuracy of biomaterial component analysis.

[0153] like Figure 3 The figure shows a functional module diagram of a biomaterial component analysis system based on the Internet of Things combined with a neural network according to the present invention.

[0154] The biomaterial composition analysis system 200 based on the Internet of Things and neural networks described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the biomaterial composition analysis system based on the Internet of Things and neural networks can include a trigger condition analysis module 201, a transportation timeliness accuracy calculation module 202, a route matching coefficient calculation module 203, a scheduling fitness analysis module 204, and a logistics route scheduling optimization module 205. The modules described in the present invention, also referred to as units, refer to a series of computer program segments that can be executed by an electronic device processor and perform a fixed function, and are stored in the electronic device's memory.

[0155] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0156] The substance content calculation module 201 is used to obtain material sample data of the biomaterial to be tested and environmental status data of the environment in which it is located, and extract the intrinsic property index of the biomaterial to be tested from the material sample data to calculate the basic substance content of the biomaterial to be tested;

[0157] The absorption intensity analysis module 202 is configured to scan the biomaterial to be tested using a preset multi-band optical probe, collect input optical signals and transmitted optical signals during the scanning process, and calculate the characteristic absorption rate of the biomaterial to be tested based on the input optical signals, the transmitted optical signals, and the basic substance content, so as to analyze the characteristic absorption intensity of the biomaterial to be tested;

[0158] The spectral information extraction module 203 is used to construct a spectral line graph corresponding to the biomaterial to be tested according to the characteristic absorption intensity, and re-extract the fingerprint spectral information of the biomaterial to be tested from the spectral line graph;

[0159] The biomaterial analysis module 204 is used to input the fingerprint spectrum information into a biological detection unit of a preset Internet of Things, use the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biomaterial to be tested, combine the molecular configuration information with the fingerprint spectrum information, and use a trained neural network to analyze the material composition of the biomaterial to be tested.

[0160] In detail, the modules in the biomaterial component analysis system 200 based on the Internet of Things combined with neural networks in the embodiment of the present invention are used in the same manner as above. Figure 1 The same technical means as the method for analyzing the composition of biological materials based on the Internet of Things combined with neural networks described in the previous section can produce the same technical effects, so I will not go into details here.

[0161] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0162] Finally, it should be noted that among the multiple embodiments described above, each embodiment can be combined with each other or be independent, and deleting any one of them will not affect the technical implementation of other embodiments. The above embodiments are only used to illustrate the technical solution of the present invention and are not limiting. Although the present invention is described in detail with reference to the preferred embodiments, ordinary technicians in this field should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for analyzing the composition of biomaterials based on the Internet of Things combined with neural networks, characterized in that: The method comprises: Obtaining material sample data of the biomaterial to be tested and environmental status data of its environment, extracting intrinsic property indicators of the biomaterial to be tested from the material sample data to calculate the basic substance content of the biomaterial to be tested; Scanning the biomaterial to be tested using a preset multi-band optical probe, collecting input optical signals and transmitted optical signals during the scanning process, and calculating the characteristic absorption rate of the biomaterial to be tested based on the input optical signal, the transmitted optical signal, and the basic substance content to analyze the characteristic absorption intensity of the biomaterial to be tested; constructing a spectrum line graph corresponding to the biomaterial to be tested according to the characteristic absorption intensity, and re-extracting fingerprint spectrum information of the biomaterial to be tested from the spectrum line graph; The fingerprint spectrum information is input into a biological detection unit of a preset Internet of Things, and the biological detection unit is used to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be tested. The molecular configuration information is combined with the fingerprint spectrum information, and the material composition of the biological material to be tested is analyzed using a trained neural network.

2. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, characterized in that: The step of extracting the intrinsic property index of the biomaterial to be tested from the material sample data includes: performing interference removal processing on the material sample data to obtain clean sample data; Identifying a feature identifier corresponding to the clean sample data, and calculating an identifier correlation degree between the feature identifier and a preset feature identifier; extracting material characteristic data from the clean sample data according to the identification correlation degree; The material characteristic data is analyzed to obtain the intrinsic property index of the biomaterial to be tested.

3. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 2, characterized in that: The calculating of the identification correlation between the feature identification and the preset feature identification includes: Performing vectorization processing on the feature identifier and the preset feature identifier respectively to obtain a first feature vector and a second feature vector; Performing dimensionality reduction processing on the first eigenvector and the second eigenvector respectively to obtain a first reduced dimensionality vector and a second reduced dimensionality vector; The first dimensionality reduction vector and the second dimensionality reduction vector are combined to calculate the identification correlation between the feature identifier and the preset feature identifier using the following formula: Among them, A represents the identification correlation between the feature identification and the preset feature identification, Represents the first dimensionality reduction vector corresponding to the ath identifier in the feature identifier, Represents the second dimensionality reduction vector corresponding to the ath identifier in the preset feature identifier, Indicates the maximum value between two values.

4. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, characterized in that: The step of extracting the intrinsic property index of the biomaterial to be tested from the material sample data to calculate the basic substance content of the biomaterial to be tested further includes: Extracting core attribute dimensions from the intrinsic attribute indicators, the core attribute dimensions including: structural feature dimensions, component feature dimensions, and environmental response feature dimensions; Standardizing the structural characteristic dimension, the component characteristic dimension, and the environmental response characteristic dimension respectively to obtain a structural standard value, a component standard value, and a response standard value; The basic substance content of the biomaterial to be tested is calculated using the following formula by combining the structure standard value, the component standard value, and the response standard value: Wherein, E represents the basic substance content of the biological material to be tested, Indicates the standard value of the structure, Indicates the standard value of the component, Indicates the response standard value, Represents the dimensionality correction factor.

5. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, characterized in that: The step of calculating the characteristic absorption rate of the biomaterial to be tested by combining the input optical signal, the transmitted optical signal, and the basic substance content includes: Measuring the detection distance between the biological material to be tested and the multi-band optical probe; determining a propagation distance of the optical signal of the biomaterial to be tested based on the detection distance; Measuring the signal intensities of the input optical signal and the transmitted optical signal respectively to obtain the input optical signal intensity and the projected optical signal intensity; The characteristic absorption rate of the biomaterial to be tested is calculated using the following formula based on the propagation distance, the input optical signal intensity, the projected optical signal intensity, and the basic substance content: in, represents the characteristic absorption rate of the biological material to be tested, d represents the propagation distance, Indicates the input optical signal strength, It represents the intensity of the projected optical signal, and E represents the basic substance content.

6. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, characterized in that: The analyzing the characteristic absorption intensity of the biomaterial to be tested further comprises: querying the probe attribute parameters of the multi-band optical probe, and determining the optical signal band of the optical signal of the biomaterial to be tested based on the probe attribute parameters; measuring the molecular vibration frequency of the biomaterial to be tested, and calculating the material characteristic coefficient of the biomaterial to be tested based on the molecular vibration frequency; The absorption contribution value of the biomaterial to be tested is calculated using the following formula in combination with the material characteristic coefficient, the optical signal band, and the characteristic absorption rate: in, Indicates the absorption contribution value of the biomaterial to be tested, represents the characteristic absorption rate, k represents the material characteristic coefficient, Indicates the optical signal band; Based on the absorption contribution value, the characteristic absorption intensity of the biological material to be tested is analyzed.

7. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, wherein: The step of extracting fingerprint spectrum information of the biological material to be tested from the spectrum line graph includes: Performing background signal removal processing on the spectral line graph to obtain a background correction spectrum line; Performing filtering on the background correction spectrum to obtain a filtered spectrum; Performing feature enhancement processing on the filtered spectral lines to obtain enhanced spectral lines; Identifying a characteristic absorption region in the enhanced spectrum and extracting an absorption mode corresponding to the characteristic absorption region; According to the absorption pattern, fingerprint spectrum information corresponding to the biological material to be tested is determined.

8. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, wherein: The extracting the absorption mode corresponding to the characteristic absorption region includes: Calculating the integrated intensity corresponding to each region in the characteristic absorption region, and determining the intensity distribution characteristics in the integrated intensity; Locating a central band corresponding to the characteristic absorption region according to the intensity distribution characteristics; Calculate the half-maximum width corresponding to each region in the characteristic absorption region, and analyze the contour shape corresponding to the characteristic absorption region; The geometric properties corresponding to the contour shape are analyzed, and the absorption pattern corresponding to the characteristic absorption region is generated by combining the central band, the half-peak width and the geometric properties.

9. The method for analyzing biomaterial components based on the Internet of Things combined with a neural network according to claim 1, wherein: The method of utilizing the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biological material to be detected includes: Analyzing data features in the fingerprint spectrum information, and configuring a combination of sensing elements in the biometric detection unit according to the data features; Adjusting the sensitivity of the sensing element combination to obtain an optimized sensing combination; Performing characteristic response processing on the fingerprint spectrum information using the optimized sensing combination to obtain a characteristic response spectrum; Converting the characteristic response spectrum into a time-series electrical signal, and generating the sensing signal sequence according to the time-series electrical signal; Response pattern features are extracted from the sensing signal sequence, and molecular configuration information of the biomaterial to be tested is identified based on the response pattern features.

10. A biomaterial component analysis system based on the Internet of Things combined with a neural network, characterized in that: The system comprises: a substance content calculation module, configured to obtain material sample data of the biomaterial to be tested and environmental status data of the environment in which it is located, extract the intrinsic property index of the biomaterial to be tested from the material sample data, and calculate the basic substance content of the biomaterial to be tested; an absorption intensity analysis module, configured to scan the biomaterial to be tested using a preset multi-band optical probe, collect input optical signals and transmitted optical signals during the scanning process, and calculate the characteristic absorption rate of the biomaterial to be tested based on the input optical signal, the transmitted optical signal, and the basic substance content, so as to analyze the characteristic absorption intensity of the biomaterial to be tested; a spectral information extraction module, configured to construct a spectral line graph corresponding to the biomaterial to be tested based on the characteristic absorption intensity, and re-extract the fingerprint spectral information of the biomaterial to be tested from the spectral line graph; The biomaterial analysis module is used to input the fingerprint spectrum information into a biological detection unit of a preset Internet of Things, use the biological detection unit to output a corresponding sensing signal sequence to identify the molecular configuration information of the biomaterial to be tested, combine the molecular configuration information with the fingerprint spectrum information, and use a trained neural network to analyze the material composition of the biomaterial to be tested.

Images